Settings + signal model for categorical pipeline

[n-gilbertson]

• Created default settings for the categorical pipeline
• Fit and plot signal model – coefficients, data, outliers.

•  (Is there a better way to get the betas, thinking particularly about the addition of covariates later on and extracting the intercept betas appropriately – using get_beta_info function from utils maybe?)
  

• Add entry point for categorical pipeline
• Set things up to ignore .DS_Store files (maybe – if implemented correctly)

[zhengp0]

here we probably can use

df[col].unique() == 1

But we can make this chance along with removing columns with only several 1s or 0s. And also in other types of pipelines

[zhengp0]

Thanks Nora!! Everything looks good to me and we can merge this PR :)